Execution algorithms (TWAP, VWAP, POV)

Execution algorithms are programs that split a large parent order into many small child orders over time to limit market impact. The common families are TWAP, which trades evenly over a period, VWAP, which follows the expected volume profile, and POV, which trades a set percentage of market volume. Their child orders leave regular, repetitive traces in the tape that order flow readers try to recognize.

Senzoukria · Glossary · Updated September 2026


The main families

Common execution algorithms
AlgorithmSlicing ruleTypical trace
TWAPEqual quantities at regular time intervalsSimilar-sized prints at a steady rhythm
VWAPQuantities proportional to the expected volume profile of the dayMore activity in the busy parts of the session
POV (participation)A fixed share of the volume that actually tradesChild orders that speed up and slow down with the market
Arrival-price / shortfallFront-loaded to limit drift from the decision priceHeavier activity early, then tapering

A worked example

A participation algorithm set at 5% trades in a market printing about 1,200 contracts a minute. It aims for roughly 60 contracts a minute, perhaps as 6 child orders of 10, some passive, some crossing the spread. A TWAP working 2,000 contracts over 100 minutes aims for 20 contracts a minute regardless of market activity. Neither order ever shows its total size; each minute looks ordinary on its own.

What they leave in the order flow

  • Repeated prints of the same size at a steady rhythm, when the algorithm crosses the spread.
  • A bid or offer that keeps refilling at the best price, when it works passively, similar to an iceberg.
  • Persistent one-sided delta with modest price impact over a long period.
  • Activity clustering around the session VWAP when the benchmark is VWAP.

The limits of detection

Recognizing an algorithm from the tape is inference. Repeated clip sizes also come from manual traders with a habit, from several unrelated participants, or from market makers requoting. Even a correctly spotted algorithm does not reveal its total size, its end time or its reason, and it can stop at any moment. Trading ahead of detected large interest is what regulators describe as order anticipation, a strategy that relies on exactly this kind of inference.

In Senzoukria

Senzoukria does not offer algorithmic execution: orders are sent from the ticket, the chart or the DOM, and the autopilot sends market entries. For reading other participants' slicing, the trades tape on the Trading page merges consecutive trades at the same price and side within 500 milliseconds and displays how many were merged, so 400 contracts in one print and 400 in fifty orders do not look alike. The indicator library adds Trade Size Distribution, Large Trade Ratio, Average Trade Size and Delta per Trade, which measure how concentrated or fragmented a bar's volume was, and the VWAP family for the benchmark itself.

In the same section

This page in other languages

Frequently asked questions

Can you detect a TWAP or VWAP order on the tape?
Sometimes, and never with certainty. Regular same-size prints or a steadily refilling best price are consistent with an algorithm, but they are also consistent with other explanations. The tape shows child orders, never the parent order's size or intention.
Why do large traders use execution algorithms?
Because a large order sent at once would consume several levels of the book and move the price against itself. Slicing it over time lets the book refill between child orders, reducing market impact at the cost of taking longer and being exposed to price drift meanwhile.

Keep reading